The
n e w e ng l a n d j o u r na l
of
m e dic i n e
Review Article Dan L. Longo, M.D., Editor
Sciatica Allan H. Ropper, M.D., and Ross D. Zafonte, D.O. From Brigham and Women’s Hospital (A.H.R.) and Spaulding Rehabilitation Hospital (R.D.Z.) — both in Boston. N Engl J Med 2015;372:1240-8. DOI: 10.1056/NEJMra1410151 Copyright © 2015 Massachusetts Medical Society.
T
he mundane malady sciatica has been known to physicians since antiquity.1 It is pain that radiates from the buttock downward along the course of the sciatic nerve,2 but the term has been used indiscriminately for a variety of back and leg symptoms. Although sciatica has several causes, Mixter and Barr extended previous observations to establish in 1934 that the principal source is compression of a lumbar nerve root by disk material that has ruptured through its surrounding annulus (see the Glossary).3 Neuroradiologic studies affirm that 85% of cases of sciatica are associated with a disk disorder.4
A nat om y of the Sci at ic Nerv e The fourth and fifth lumbar nerve roots and the first two sacral nerve roots join in the lumbosacral plexus to form the peroneal and tibial nerves that leave the pelvis in an ensheathed single trunk as the sciatic nerve, the largest nerve in the body. Disturbances anywhere along the course of the sciatic nerve can give rise to sciatica, but the most common areas are at the sites of disk rupture and osteoarthritic change — at the L4–L5 and L5–S1 levels and, less frequently, the L3–L4 level — where there is usually compression of the root below the corresponding disk. The mechanism of sciatica may relate to distortion of the nerve root or its sensory ganglion5 or putatively to the effect of local inflammatory cytokines. Other sites of sciatic injury are in the lower pelvic cavity, buttock, gluteal fold, and proximal biceps femoris muscle (Fig. 1). The prevalence of sciatica varies widely among studies, with a highest incidence of 40%. Most instances occur in the fourth and fifth decades of life.6
S ymp t oms a nd E x a minat ion Findings Sciatica may begin either suddenly with physical activity or slowly. Sciatic pain has aching and sharp components and radiates along a broad line from the middle or lower buttock, proceeding dorsolaterally in the thigh in cases of compression of the L5 nerve root and posteriorly in cases of compression of S1. With L4 compression, the pain is anterolateral in the thigh and may be misattributed to hip disease. If nerve-root pain extends below the knee, its location conforms to the superficial sensory distribution of the affected spinal root. Sciatica is usually unilateral, in keeping with the common dorsolateral configuration of disk rupture and with foraminal stenosis from osteoarthritic disease of the spine. There may be bilateral pain in central disk herniation, lumbar stenosis, and spondylolisthesis. Low back pain of variable severity accompanies sciatica but is not a consistent feature. Aching in the L5–S1 area or in the upper sacroiliac joint is common with disk rupture. Increased back and sciatic pain with coughing, sneezing, straining, or other forms of the Valsalva maneuver suggests disk rupture. Depending on the site 1240
n engl j med 372;13 nejm.org March 26, 2015
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
Sciatica
Glossary Annulus: The annulus is the connective tissue surrounding the spinal disk. Biceps femoris: This muscle is also known as the hamstring muscle. Fibrillations: Fibrillations are spontaneous short-duration electrical potentials of single muscle fibers seen on electromyography and caused by denervation of muscle. They are not visible by observing the surface of the skin. Foraminal stenosis: This term refers to narrowing of the canal of exit of spinal nerve roots. Late responses (F-wave and H-reflex): The F-wave is an elicited electrical response of the anterior horn cell to supramaximal stimulation of a motor nerve. An antidromic stimulus of the motor axon depolarizes the anterior horn cell and causes an orthodromic potential that appears as a second motor action potential after the initial compound motor action potential. The H-reflex is the electrical representation of the ankle reflex. The F-wave and the H-reflex are useful because they are conducted through the spinal nerve root and proximal nerve and are therefore sensitive to compression by a disk. Nocifensive response: This response is an involuntary, defensive muscular reaction that prevents use or movement of a painful body part. Sciatic notch (greater sciatic notch): The sciatic notch is the point of exit of the sciatic nerve under the ischial spine at the pelvic brim. Sharp waves: Sharp waves are similar to fibrillations but appear earlier. Both types of electrical potentials take weeks or longer to appear after damage to a motor axon. Spondylolisthesis: Spondylolisthesis is displacement (subluxation) of a vertebral body and its posterior elements from its normal alignment.
and size of the rupture, the patient may adopt a posture of ventroflexion and either reduced or exaggerated lumbar lordosis, minimizing pressure on the root. Bilateral sciatica that is brought on by walking and simulates vascular claudication is the result of compression of the cauda equina roots. Such compression is known as neurogenic claudication (the Verbiest syndrome). Patients may have paresthesias in the dermatomal distribution of a nerve root, but sensory symptoms and signs are not prominent. Weakness is present in less than half of patients but is infrequently severe enough to cause foot drop (in cases of L5 radiculopathy) or a downward tilted pelvis during walking (in cases of gluteal weakness due to compression of S1). Compression of the S1 nerve root is usually associated with reduction in or loss of the ankle reflex, and L3 or L4 compression is associated with variable reduction in the knee reflex; L5 compression causes inconsistent changes in reflexes. There are frequently unanticipated variations of these reflex patterns. Many clinical tests have been devised to determine whether sciatic pain is caused by disk compression of a spinal nerve root; most of the tests are variations of the straight-leg-raising test (Lasègue’s test) (Fig. 2). In a patient in the supine position, raising the leg with the knee extended stretches the nerve root over the protruded disk and results in a nocifensive response of muscle
contraction. A positive test consists of reproduction or marked worsening of the patient’s initial pain and firm resistance to further elevation of the leg. A diagnosis of disk compression is likely if pain radiates from the buttock to below the knee when the angle of the leg is between 30 and 70 degrees. Sensitivity of the test for disk herniation is approximately 90%, but specificity is low.7 Many persons without spinal abnormalities have hamstring and gluteal tightness with discomfort elicited by straight-leg raising, but the pain is more diffuse than in sciatica and the leg can be lifted higher if the maneuver is performed slowly. Increased pain on dorsiflexion of the foot or large toe increases sensitivity. The crossed straight-leg-raising test (Fajersztajn’s test) involves raising the unaffected leg; in a positive test, sciatic pain is elicited in the opposite (affected) leg. This test is 90% specific for disk herniation on the contralateral side but is insensitive.7
Im aging a nd El ec t roph ysiol o gic a l Te s t ing Imaging and electrophysiological evidence of nerve-root compression corroborates structural disk or spine disease as the proximate cause of sciatica, but testing is not necessary in a typical case until intervention is required. Radiographs of the lumbar spine provide limited information, but they may show reduction in the height of an
n engl j med 372;13 nejm.org March 26, 2015
1241
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
A
Normal Anatomy
The
n e w e ng l a n d j o u r na l
B
Disk Herniation
of
m e dic i n e
C
L1 ANTERIOR VIEW
Spondylolisthesis
Displacement of the lumbar vertebra associated with degenerative spine disease or trauma
Results from an annulus fibrosus tear, allowing the nucleus pulposus to herniate and compress the nerve root exiting one level below
Path of spinal nerves
L2 Disk herniation at L4–L5
Cauda equina L3
L4
L4 LATERAL
L4 root
L4 Greater sciatic foramen
L5 root
L5
Intervertebral disk L5–S1
Compression of nerve root L5
Lumbosacral plexus
VIEW
L5
L5
Anterior Sacrum slippage of lumbar vertebrae
POSTERIOR VIEW
D
Foraminal Stenosis
E
Apophyseal joint hypertophy Intervertebral disk Foraminal stenosis
Sacrum
Tibial nerve
F
Peroneal nerve
Synovial cyst Apophyseal joint Cyst develops as a result of apophyseal joint degeneration
Ligamentous hypertrophy
Spinal nerve
Sciatic nerve
Synovial Cyst
Spinal nerves
G
Piriformis Syndrome
Gluteal Injection-Site Trauma
Injury of the piriformis muscle compresses the sciatic nerve Sacrum
Femur
Piriformis muscle
Sciatic nerve
POSTERIOR VIEW
H
Gluteus maximus
Obstetrical Sciatic Compression
I
Pubis
Head of fetus compresses sciatic nerve against the sacrum
Pelvic-Floor Tumors Abdominal cavity
Birth canal
Sciatic nerve
Injury varies depending on the location and agent being injected
Tumor
Soft-tissue tumor exerts pressure on the lumbosacral plexus and sciatic nerve
Head of fetus Lumbosacral plexus
Sciatic nerve
INFERIOR
Sacrum
Pelvis
Sciatic nerve
L ATERAL
VIEW
VIEW
Figure 1. Origin and Course of the Sciatic Nerve and the Main Sites of Damage. Panel A shows the normal anatomy surrounding the sciatic nerve. Panels B through I show the main types of damage.
1242
n engl j med 372;13
nejm.org
March 26, 2015
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
Sciatica
A
Lasègue’s Test
Increased pain on dorsiflexion of the patient’s foot increases sensitivity of the test Peroneal nerve Tibial nerve
Sciatic nerve
40 to 50 degrees
B
Spinal Nerve Roots under Traction
L5 nerve root
Dorsal root ganglion
Intervertebral disk L3–L4
Cauda equina L4
L5
L3
Pedicle
L2
Sacrum
Herniated disk at L4–L5 compressing L5 nerve root POSTERIOR VIEW
Figure 2. Straight-Leg-Raising Test. Panel A shows the straight-leg-raising test, with assessment for the presence or absence of Lasègue’s sign, and Panel B shows the spinal nerve roots under traction during the test.
intervertebral space, spondylolisthesis, osteomyelitis, or tumor infiltration of a vertebral body. The nature and location of disk rupture and spinal lesions, such as osteoarthritic disease and spondylolisthesis, lateral recess stenosis, and synovial cysts of the facet joint, can be accurately seen on magnetic resonance imaging (MRI) without the administration of gadolinium (Fig. 3). n engl j med 372;13
Computed tomography (CT) is performed less frequently but reveals most disk herniations and structural changes of the spine. CT and MRI each have a role in exploration of the pelvic fossa when sciatic compression is suspected in those regions. Disk bulging and related minor deformities generally do not cause compression of nerve roots and are not an explanation for persistent sciati-
nejm.org
March 26, 2015
1243
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
The
n e w e ng l a n d j o u r na l
of
m e dic i n e
findings are fibrillations and sharp waves in muscles corresponding to a single nerve root, combined with normal sensory potentials from nerves that carry that root; the latter is the result of the preganglionic site of root compression, leaving the cell body and distal axon unaffected. Sampling four or five relevant muscles including the paraspinal regions is considered adequate.9 Late responses are often included in the nerveconduction studies because they are sensitive to root compression. Electrodiagnostic confirmation that sciatica is due to a radiculopathy rather than to an undefined musculoskeletal problem is associated with a better outcome,10 but the role of EMG in sciatica has not been established and some guidelines do not require testing.
A
Nonspina l C ause s of Sci at ic a
B
Sciatica due to disk rupture and degenerative spine disease is more common than sciatica due to all nonspinal causes taken together (Table 1 and Fig. 1). Hip disease and sciatica are likely to be mistaken for each other. Piriformis Syndrome and Back-Pocket Sciatica
Figure 3. MRI Scan of a Ruptured Intervertebral Disk. Sagittal (Panel A) and axial (Panel B) T2 -weighted images show a herniated disk (arrows) at L4–L5. Compression of the right L5 nerve root can be seen as it exits the neural foramen (arrowhead).
ca. The clinician should be aware that a high rate of minor abnormalities are seen on MRI in patients without back or radicular symptoms but that disk rupture is seen in less than 1% of asymptomatic patients.8 Needle electromyography (EMG) and nerveconduction studies aid diagnosis by revealing a topographic distribution of muscular denervation corresponding to a nerve root. Denervation becomes apparent days or weeks after injury, and the appearance of changes takes longer in distal than in proximal muscles. The most specific
1244
n engl j med 372;13
The piriformis syndrome has been attributed to compression of the sciatic nerve underlying the piriformis muscle, which stabilizes and serves as an external rotator of the hip. Although the nature and frequency of the syndrome are uncertain, the most consistent features, derived from a systematic review,11 are focal mid-buttock pain, tenderness over the sciatic notch, aggravation of pain after sitting, and provocation with maneuvers that increase piriformis muscle tension such as external rotation of the hip. Pain with the straightleg-raising maneuver was reported in a metaanalysis to occur in approximately half of cases. It is presumed that an injured piriformis muscle, sometimes caused by running, stretching, or lunging, compresses the sciatic nerve, but the pathophysiology is unclear. Electrodiagnostic and imaging tests are usually normal. In the mid-20th century, as wallets expanded in capacity, sciatica (“credit-carditis”) that was cured by removal of the wallet from the back pocket became the subject of anecdotal reports.12,13 Tools, cell phones, and golf balls in back pockets and prolonged sitting on hard surfaces, including car seats, can also produce sciatica.
nejm.org
March 26, 2015
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
Sciatica
Zoster Sine Herpete
For the first few days before the eruption of shingles in a lumbar or upper sacral dermatome, herpes zoster reactivation closely simulates sciatica caused by disk rupture. Diagnosis is difficult in the intervening days and especially in the few instances in which the rash never develops.
Table 1. Causes of Sciatica and of Conditions Simulating Sciatica. Spinal causes Disk rupture with compression of L4, L5, or S1 root Osteoarthritic impingement on roots due to spondylolisthesis or foraminal or lumbar stenosis Synovial cyst of facet joint
Traumatic Injury of the Sciatic Nerve
Arachnoid (Tarlov) cyst
Traumatic injury of the sciatic nerve occurs with fractures of the pelvis or after proximal hamstring injury and with extreme stretching of the nerve. Muscle hematomas or tendinous injuries may cause severe sciatica. Posterior hip dislocation or femoral fracture impinge on the nerve; sciatica may also occur during relocation of the hip joint. Injection-site damage in the buttocks is a self-evident but infrequent cause.
Tumors of the spinal column
Gynecologic and Peripartum Causes
Zoster sine herpete
Neurofibromas on lumbar and sacral nerve roots Arachnoiditis Nonspinal causes Pelvic and gynecologic conditions, including cyclic sciatica due to endometriosis Piriformis syndrome and back-pocket sciatica Pregnancy, delivery, and prolonged time in the lithotomy position
Diabetic radiculopathy Deposits of endometrial tissue on the proximal Gluteal injection-site trauma nerve can cause recurrent cyclic, or catamenial, sciatica, more often on the right side than on the Lumbar plexitis left side.14 Large ovarian cysts and uterine enlargeVascular impingement on sciatic nerve (pseudoment associated with late pregnancy can compress aneurysm of the inferior gluteal artery) the nerve between the fetal head and the pelvic Hip fracture and dislocation or relocation; extensive pelvic fractures brim (Fig. 1). It is not unusual for a woman to have unilateral or bilateral sciatica post partum, Biceps femoris hematoma, strain, and tear regardless of whether forceps were used during Idiopathic causes in young persons delivery; sciatica ostensibly occurs after a woman has spent a prolonged time in the lithotomy position. tions may provide short-term relief for low back and sciatic pain; however, it is difficult to determine their effect on sciatica,15 and many patients C onservat i v e T r e atmen t report little relief. Orally or systemically adminof Sci at ic a istered glucocorticoids have been used to amelioSciatica resolves without treatment in the major- rate sciatica, but it is difficult to interpret their ity of cases. Numerous systematic analyses have effect.15-17 Many guidelines recommend restrictions compared various forms of therapy for sciatica, but on the use of opioids. Antiepileptic drugs, gabathe validity of their constituent studies is limit- pentin, pregabalin, antidepressant agents (e.g., tried.15-17 Many conflate treatment for back pain with cyclic agents), muscle relaxants, and pain medicathat for sciatica. Although it is difficult to derive tions have been used but with little supporting data. clinically useful information from these compiAny intervention that temporarily reduces pain lations, they do provide some perspective on the and increases participation in exercises, even if it relative value of treatments. is associated with no long-term benefit, could play The most common initial treatment is pain a role in care. The benefits of physical therapy and control by means of medication and physical various exercise regimens are difficult to detertherapy. Activity is usually self-limited in propor- mine, and the superiority of any one program has tion to the degree of discomfort, and although not been established, although most appear to be rest is often recommended, it is not better than safe. Programs include directional preference exmovement in patients who are able to remain ercises (back-movement exercises in a direction active.18 Nonsteroidal antiinflammatory medica- that causes the locus of pain to move proximally,
n engl j med 372;13 nejm.org March 26, 2015
1245
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
The
n e w e ng l a n d j o u r na l
toward the mid-back, where it is better tolerated),19 motor-control exercises (also known as specific stabilization exercises) that focus on enhancing control of the transversus abdominis and multifidus muscles, which stabilize the spine, strengthening of other core muscles, stretching, general fitness exercises, and yoga. One conventional approach involves slowly increasing mobilization of the lumbar spinal segments by means of stretching and exercise, improving posture, and strengthening the muscles that stabilize the spinal column and pelvis. It has been difficult to show that this approach accelerates recovery or prevents future injury, but some trials suggest that it is superior to rest in the acute phase of sciatica.20 Spinal manipulation for sciatica is widely used and has been studied with an assortment of designs and comparators; therefore, reviews of existing trials, most considered of low or moderate quality, draw limited conclusions.16,17,21 One recent pragmatic trial for subacute or chronic back and leg pain has suggested that spinal manipulation provides a short-term benefit when added to an exercise program.22 Lumbar traction was used in the past, particularly if low back pain was prominent, but a review of available studies, most of low and moderate quality, showed no advantage.23 Transcutaneous electrical nerve stimulation is probably ineffective.24 Acupuncture has been proposed for persistent sciatic discomfort, but its efficacy has not been shown.15,16 Chemonucleolysis of an extruded disk had positive results in some studies and systematic analyses but is infrequently used. Epidural injections of glucocorticoids are frequently administered for low back pain and related conditions. Trials have suggested an associated short-term decrease in leg pain but no decrease in the need for subsequent surgery.16,17,25 There is no clear difference in efficacy between fluoroscopically guided translaminar and transforaminal injections, but some guidelines endorse one technique over the other. In experimental studies and preliminary clinical trials, biologic agents such as tumor necrosis factor inhibitors have shown negative or limited positive effects on leg pain.26 Treatment of the piriformis syndrome involves stretching and physical therapy to enhance mobility. Injections into the muscle are sometimes administered under fluoroscopic, electromyographic, or ultrasonographic guidance; the use of glu1246
of
m e dic i n e
cocorticoids and botulinum toxin has been examined alone and in combination with physical therapy, with the suggestion that botulinum toxin may be more helpful than either placebo or injections containing a combination of lidocaine and a glucocorticoid in relieving discomfort,27 but support for any of these treatments is weak.
Surgic a l T r e atmen t of Sci at ic a C aused by Lumb a r Disk Dise a se Sciatica resolves without treatment in one third of patients within 2 weeks and in three quarters of patients within 3 months after onset.28 Nevertheless, most trials comparing surgical treatment and conservative treatment of sciatica due to lumbar disk disease favor surgery, because it results in earlier relief of pain. One representative study showed that patients who had sciatica for 6 to 12 weeks and were assigned to lumbar disk surgery had faster and more pronounced pain relief than those assigned to conservative treatment. However, 1 year after surgery, there were few differences in pain or disability between the two groups.29 The trial also showed that 39% of patients initially assigned to conservative therapy required surgery after a median of 14 weeks of pain, and in a 5-year follow-up study, an additional 7% requested surgery.30 One may conclude that surgery produces faster pain relief and accelerates improvement in mobility but that postponement of surgery to determine whether pain will abate is appropriate. It is not known whether delaying surgery in persons with weakness reduces the chance of full recovery or allows time for weakness to resolve. Surgery has been recommended in cases of large disk ruptures into the spinal canal that compress the cauda equina and cause failure of the bladder sphincter or the bowel sphincter. In a cost-effectiveness study of prolonged conservative care versus early surgery, the shortened duration of pain after diskectomy correlated with a cost savings per quality-adjusted life-year of approximately $60,000.31 In that study, 23% of the participants reverted from recovered to not recovered or vice versa at various times over a 5-year period, which indicates that sciatica can be a chronic and relapsing symptom. In another study, the configuration of the disk on MRI 1 year after the onset of symptoms did not distinguish between patients in whom sciatica had improved and those in whom it had not improved; inexpli-
n engl j med 372;13 nejm.org March 26, 2015
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
Sciatica
cably, this was true for patients who underwent treatment of low back pain in patients with and surgical treatment and for those who underwent in those without sciatica, the outcomes and the conservative treatment.32 cost of care have remained unchanged for more than a decade.35 A review of major trials with adequate data for analysis concluded that there Surgic a l Technique s was conflicting evidence on long-term benefit but Treatment of sciatica by means of decompression that surgery relieved pain more rapidly and to a of a lumbar nerve root is most likely to succeed if greater degree than did conservative therapy.36 symptoms conform to the typical clinical pattern A discursive analysis by the National Institute and imaging studies show disk rupture. Several of Health Research, in the United Kingdom, about open and percutaneous surgical approaches have the effectiveness of various strategies to manage been devised. A unilateral hemilaminotomy (re- sciatica15,37 showed some support for almost all moving parts of adjacent laminae on the side that current treatments, including disk surgery, epineeds to be decompressed) is usually adequate. dural glucocorticoids, chemonucleolysis, and alIn the past, bilateral laminectomy, a more exten- ternative therapies. However, only surgery bensive procedure, was performed, but the unilateral efitted all aspects of global effect, pain relief, and procedure is favored because it preserves tension a composite of condition-specific outcomes in and alignment between adjacent spinal segments. the short, medium, and long term. Microdiskectomy and various minimally invasive North American Spine Society guidelines38 and percutaneous techniques involving, for ex- state that diskectomy provides more effective and ample, a series of tubular retractors or an endo- more rapid symptom relief than do other treatscope, are also used to treat disk rupture. ments for symptoms that warrant surgery, alThe sequestered disk fragment or larger though less severe symptoms can be managed amounts of the intra-annular disk may be re- conservatively; patients with psychological dismoved. Fusion of adjacent spinal segments is gen- tress have poor outcomes after surgery; epidural erally not necessary if there is no mobile spondy- glucocorticoid injection provides short-term pain lolisthesis and the operation is at one level. In relief; electrodiagnostic studies have limited utilpooled analyses, there is little evidence that any ity; and for many technical aspects of surgery, technique, minimally invasive or conventional, such as lumbar fusion, there was insufficient produces better results than any other technique.33 evidence for comment. The guidelines also indiMinimally invasive approaches show a trend to- cate that there is inadequate evidence to gauge ward earlier and better pain relief, probably be- how long one can wait before surgery and still cause there is less paraspinal muscle damage, but recover from cauda equina compression or moaccording to a meta-analysis, they require longer tor deficits. British Pain Society guidelines for the treatoperating times and are associated with a higher rate of rerupture of the disk34 and therefore may ment of low back pain and radicular pain39 broadnot be appropriate in all cases. Complications of ly reflect other analyses in separating the plans for surgery are infrequent but include dural tears radicular pain from those for low back pain and with leakage of cerebrospinal fluid, as well as acknowledge that the recommendations lack predamage to the root or cauda equina. The use of cision. According to the British Pain Society, any hardware to prevent instability and chronic low indication of cauda equina compression requires back pain is often necessary in cases of spinal referral for urgent surgery. Severe radicular pain stenosis, in which multilevel operations or those that is disabling or intrusive or that prevents the involving wide resection of bone are typically patient from performing everyday tasks requires performed. MRI, as does a persistent neurologic deficit that lasts 2 weeks. MRI should be ordered by clinicians who can interpret the results; it is discourGuidel ine s a nd S ys tem at ic aged at the primary care level. R e v ie ws Despite the numerous guidelines and systematic reviews that have been published regarding the
No potential conflict of interest relevant to this article was reported. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.
n engl j med 372;13 nejm.org March 26, 2015
1247
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
Sciatica
References 1. Pearce JMS. A brief history of sciatica. 16. Luijsterburg PA, Verhagen AP, Ostelo Spinal Cord 2007;45:592-6. RW, van Os TA, Peul WC, Koes BW. Ef2. Frymoyer JW. Back pain and sciatica. fectiveness of conservative treatments for N Engl J Med 1988;318:291-300. the lumbosacral radicular syndrome: a 3. Mixter WJ, Barr JS. Rupture of the in- systematic review. Eur Spine J 2007; 16: tervertebral disk with involvement of the 881-99. spinal canal. N Engl J Med 1934;211:210-5. 17. Valat JP, Genevay S, Marty M, Rozen4. Porchet F, Wietlisbach V, Burnand B, berg S, Koes B. Sciatica. Best Pract Res Daeppen K, Villemure JG, Vader JP. Rela- Clin Rheumatol 2010;24:241-52. tionship between severity of lumbar disc 18. Vroomen PC, de Krom MC, Wilmink disease and disability scores in sciatica JT, Kester AD, Knottnerus JA. Lack of efpatients. Neurosurgery 2002;50:1253-9. fectiveness of bed rest for sciatica. N Engl 5. Lindblom K, Rexed B. Spinal nerve J Med 1999;340:418-23. injury in dorso-lateral protrusions of lum- 19. Surkitt LD, Ford JJ, Hahne AJ, Pizzari bar discs. Neurosurgery 1948;5:413-32. T, McMeeken JM. Efficacy of directional 6. Konstantinou K, Dunn KM. Sciatica: preference management for low back review of epidemiological studies and pain: a systematic review. Phys Ther 2012; prevalence estimates. Spine (Phila Pa 1976) 92:652-65. 2008;33:2464-72. 20. Kennedy DJ, Noh MY. The role of core 7. van der Windt DA, Simons E, Ripha- stabilization in lumbosacral radiculopagen II, et al. Physical examination for thy. Phys Med Rehabil Clin N Am 2011;22: lumbar radiculopathy due to disc hernia- 91-103. tion in patients with low-back pain. Co- 21. Leininger B, Bronfort G, Evans R, Rechrane Database Syst Rev 2010; 2: iter T. Spinal manipulation or mobilizaCD007431. tion for radiculopathy: a systematic re8. Jensen MC, Brant-Zawadzki MN, Obu- view. Phys Med Rehabil Clin N Am 2011; chowski N, Modic MT, Malkasian D, Ross 22:105-25. JS. Magnetic resonance imaging of the 22. Bronfort G, Hondras MA, Schulz CA, lumbar spine in people without back pain. Evans RL, Long CR, Grimm R. Spinal maN Engl J Med 1994;331:69-73. nipulation and home exercise with advice 9. Tsao BE, Levin KH, Bodner RA. Com- for subacute and chronic back-related leg parison of surgical and electrodiagnostic pain: a trial with adaptive allocation. Ann findings in single root lumbosacral ra- Intern Med 2014;161:381-91. diculopathies. Muscle Nerve 2003;27:60- 23. Wegner I, Widyahening IS, van Tulder 4. MW, et al. Traction for low-back pain with 10. Savage NJ, Fritz JM, Kircher JC, Thac- or without sciatica. Cochrane Database keray A. The prognostic value of electro- Syst Rev 2013;8:CD003010. diagnostic testing in patients with sciati- 24. Buchmuller A, Navez M, Milletre-Berca receiving physical therapy. Eur Spine J nardin M, et al. Value of TENS for relief of 2014 July 22 (Epub ahead of print). chronic low back pain with or without 11. Hopayian K, Song F, Riera R, Sam- radicular pain. Eur J Pain 2012;16:656-65. bandan S. The clinical features of the 25. Carette S, Leclaire R, Marcoux S, et al. piriformis syndrome: a systematic review. Epidural corticosteroid injections for sciEur Spine J 2010;19:2095-109. atica due to herniated nucleus pulposus. 12. Battle JD. “Credit-carditis” — a new N Engl J Med 1997;336:1634-40. clinical entity? N Engl J Med 1966;274: 26. Genevay S, Viatte S, Finckh A, Zuffer467. ey P, Balagué F, Gabay C. Adalimumab in 13. Gould N. Back-pocket sciatica. N Engl severe and acute sciatica: a multicenter, J Med 1974;290:633. randomized, double-blind, placebo-con14. Forrest JS, Brooks DL. Cyclic sciatica trolled trial. Arthritis Rheum 2010; 62: of endometriosis. JAMA 1972;222:1177-8. 2339-46. 15. Lewis RA, Williams NH, Sutton AJ, et 27. Fishman LM, Anderson C, Rosner B. al. Comparative clinical effectiveness of BOTOX and physical therapy in the treatmanagement strategies for sciatica: sys- ment of piriformis syndrome. Am J Phys tematic review and network meta-analy- Med Rehabil 2002;81:936-42. ses. Spine J 2013 October 4 (Epub ahead 28. Vroomen PC, de Krom MC, of print). Knottnerus JA. Predicting the outcome of
1248
sciatica at short-term follow-up. Br J Gen Pract 2002;52:119-23. 29. Peul WC, van Houwelingen HC, van den Hout WB, et al. Surgery versus prolonged conservative treatment for sciatica. N Engl J Med 2007;356:2245-56. 30. Lequin MB, Verbaan D, Jacobs WCH, et al. Surgery versus prolonged conservative treatment for sciatica: 5-year results of a randomised controlled trial. BMJ Open 2013;3(5):e002534. 31. van den Hout WB, Peul WC, Koes BW, Brand R, Kievit J, Thomeer RT. Prolonged conservative care versus early surgery in patients with sciatica from lumbar disc herniation: cost utility analysis alongside a randomised controlled trial. BMJ 2008; 336:1351-4. 32. el Barzouhi A, Vleggeert-Lankamp CL, Lycklama à Nijeholt GJ, et al. Magnetic resonance imaging in follow-up assessment of sciatica. N Engl J Med 2013; 368:999-1007. 33. Jacobs WC, Arts MP, van Tulder MW, et al. Surgical techniques for sciatica due to herniated disc, a systematic review. Eur Spine J 2012;21:2232-51. 34. Chang X, Chen B, Li HY, Han XB, Zhou Y, Li CQ. The safety and efficacy of minimally invasive discectomy: a metaanalysis of prospective randomised controlled trials. Int Orthop 2014;38:1225-34. 35. Mafi JN, McCarthy EP, Davis RB, Landon BE. Worsening trends in the management and treatment of back pain. JAMA Intern Med 2013;173:1573-81. 36. Jacobs WC, van Tulder M, Arts M, et al. Surgery versus conservative management of sciatica due to a lumbar herniated disc: a systematic review. Eur Spine J 2011; 20:513-22. 37. Lewis R, Williams N, Matar HE, et al. The clinical effectiveness and cost-effectiveness of management strategies for sciatica: systematic review and economic model. Health Technol Assess 2011; 15: 1-578. 38. Kreiner DS, Hwang SW, Easa JE, et al. An evidence-based clinical guideline for the diagnosis and treatment of lumbar disc herniation with radiculopathy. Spine J 2014;14:180-91. 39. Lee J, Gupta S, Price C, Baranowski AP Low back and radicular pain: a pathway for care developed by the British Pain Society. Br J Anaesth 2013;111:112-20. Copyright © 2015 Massachusetts Medical Society.
n engl j med 372;13 nejm.org March 26, 2015
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
n e w e ng l a n d j o u r na l
of
m e dic i n e
Review Article Dan L. Longo, M.D., Editor
Sciatica Allan H. Ropper, M.D., and Ross D. Zafonte, D.O. From Brigham and Women’s Hospital (A.H.R.) and Spaulding Rehabilitation Hospital (R.D.Z.) — both in Boston. N Engl J Med 2015;372:1240-8. DOI: 10.1056/NEJMra1410151 Copyright © 2015 Massachusetts Medical Society.
T
he mundane malady sciatica has been known to physicians since antiquity.1 It is pain that radiates from the buttock downward along the course of the sciatic nerve,2 but the term has been used indiscriminately for a variety of back and leg symptoms. Although sciatica has several causes, Mixter and Barr extended previous observations to establish in 1934 that the principal source is compression of a lumbar nerve root by disk material that has ruptured through its surrounding annulus (see the Glossary).3 Neuroradiologic studies affirm that 85% of cases of sciatica are associated with a disk disorder.4
A nat om y of the Sci at ic Nerv e The fourth and fifth lumbar nerve roots and the first two sacral nerve roots join in the lumbosacral plexus to form the peroneal and tibial nerves that leave the pelvis in an ensheathed single trunk as the sciatic nerve, the largest nerve in the body. Disturbances anywhere along the course of the sciatic nerve can give rise to sciatica, but the most common areas are at the sites of disk rupture and osteoarthritic change — at the L4–L5 and L5–S1 levels and, less frequently, the L3–L4 level — where there is usually compression of the root below the corresponding disk. The mechanism of sciatica may relate to distortion of the nerve root or its sensory ganglion5 or putatively to the effect of local inflammatory cytokines. Other sites of sciatic injury are in the lower pelvic cavity, buttock, gluteal fold, and proximal biceps femoris muscle (Fig. 1). The prevalence of sciatica varies widely among studies, with a highest incidence of 40%. Most instances occur in the fourth and fifth decades of life.6
S ymp t oms a nd E x a minat ion Findings Sciatica may begin either suddenly with physical activity or slowly. Sciatic pain has aching and sharp components and radiates along a broad line from the middle or lower buttock, proceeding dorsolaterally in the thigh in cases of compression of the L5 nerve root and posteriorly in cases of compression of S1. With L4 compression, the pain is anterolateral in the thigh and may be misattributed to hip disease. If nerve-root pain extends below the knee, its location conforms to the superficial sensory distribution of the affected spinal root. Sciatica is usually unilateral, in keeping with the common dorsolateral configuration of disk rupture and with foraminal stenosis from osteoarthritic disease of the spine. There may be bilateral pain in central disk herniation, lumbar stenosis, and spondylolisthesis. Low back pain of variable severity accompanies sciatica but is not a consistent feature. Aching in the L5–S1 area or in the upper sacroiliac joint is common with disk rupture. Increased back and sciatic pain with coughing, sneezing, straining, or other forms of the Valsalva maneuver suggests disk rupture. Depending on the site 1240
n engl j med 372;13 nejm.org March 26, 2015
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
Sciatica
Glossary Annulus: The annulus is the connective tissue surrounding the spinal disk. Biceps femoris: This muscle is also known as the hamstring muscle. Fibrillations: Fibrillations are spontaneous short-duration electrical potentials of single muscle fibers seen on electromyography and caused by denervation of muscle. They are not visible by observing the surface of the skin. Foraminal stenosis: This term refers to narrowing of the canal of exit of spinal nerve roots. Late responses (F-wave and H-reflex): The F-wave is an elicited electrical response of the anterior horn cell to supramaximal stimulation of a motor nerve. An antidromic stimulus of the motor axon depolarizes the anterior horn cell and causes an orthodromic potential that appears as a second motor action potential after the initial compound motor action potential. The H-reflex is the electrical representation of the ankle reflex. The F-wave and the H-reflex are useful because they are conducted through the spinal nerve root and proximal nerve and are therefore sensitive to compression by a disk. Nocifensive response: This response is an involuntary, defensive muscular reaction that prevents use or movement of a painful body part. Sciatic notch (greater sciatic notch): The sciatic notch is the point of exit of the sciatic nerve under the ischial spine at the pelvic brim. Sharp waves: Sharp waves are similar to fibrillations but appear earlier. Both types of electrical potentials take weeks or longer to appear after damage to a motor axon. Spondylolisthesis: Spondylolisthesis is displacement (subluxation) of a vertebral body and its posterior elements from its normal alignment.
and size of the rupture, the patient may adopt a posture of ventroflexion and either reduced or exaggerated lumbar lordosis, minimizing pressure on the root. Bilateral sciatica that is brought on by walking and simulates vascular claudication is the result of compression of the cauda equina roots. Such compression is known as neurogenic claudication (the Verbiest syndrome). Patients may have paresthesias in the dermatomal distribution of a nerve root, but sensory symptoms and signs are not prominent. Weakness is present in less than half of patients but is infrequently severe enough to cause foot drop (in cases of L5 radiculopathy) or a downward tilted pelvis during walking (in cases of gluteal weakness due to compression of S1). Compression of the S1 nerve root is usually associated with reduction in or loss of the ankle reflex, and L3 or L4 compression is associated with variable reduction in the knee reflex; L5 compression causes inconsistent changes in reflexes. There are frequently unanticipated variations of these reflex patterns. Many clinical tests have been devised to determine whether sciatic pain is caused by disk compression of a spinal nerve root; most of the tests are variations of the straight-leg-raising test (Lasègue’s test) (Fig. 2). In a patient in the supine position, raising the leg with the knee extended stretches the nerve root over the protruded disk and results in a nocifensive response of muscle
contraction. A positive test consists of reproduction or marked worsening of the patient’s initial pain and firm resistance to further elevation of the leg. A diagnosis of disk compression is likely if pain radiates from the buttock to below the knee when the angle of the leg is between 30 and 70 degrees. Sensitivity of the test for disk herniation is approximately 90%, but specificity is low.7 Many persons without spinal abnormalities have hamstring and gluteal tightness with discomfort elicited by straight-leg raising, but the pain is more diffuse than in sciatica and the leg can be lifted higher if the maneuver is performed slowly. Increased pain on dorsiflexion of the foot or large toe increases sensitivity. The crossed straight-leg-raising test (Fajersztajn’s test) involves raising the unaffected leg; in a positive test, sciatic pain is elicited in the opposite (affected) leg. This test is 90% specific for disk herniation on the contralateral side but is insensitive.7
Im aging a nd El ec t roph ysiol o gic a l Te s t ing Imaging and electrophysiological evidence of nerve-root compression corroborates structural disk or spine disease as the proximate cause of sciatica, but testing is not necessary in a typical case until intervention is required. Radiographs of the lumbar spine provide limited information, but they may show reduction in the height of an
n engl j med 372;13 nejm.org March 26, 2015
1241
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
A
Normal Anatomy
The
n e w e ng l a n d j o u r na l
B
Disk Herniation
of
m e dic i n e
C
L1 ANTERIOR VIEW
Spondylolisthesis
Displacement of the lumbar vertebra associated with degenerative spine disease or trauma
Results from an annulus fibrosus tear, allowing the nucleus pulposus to herniate and compress the nerve root exiting one level below
Path of spinal nerves
L2 Disk herniation at L4–L5
Cauda equina L3
L4
L4 LATERAL
L4 root
L4 Greater sciatic foramen
L5 root
L5
Intervertebral disk L5–S1
Compression of nerve root L5
Lumbosacral plexus
VIEW
L5
L5
Anterior Sacrum slippage of lumbar vertebrae
POSTERIOR VIEW
D
Foraminal Stenosis
E
Apophyseal joint hypertophy Intervertebral disk Foraminal stenosis
Sacrum
Tibial nerve
F
Peroneal nerve
Synovial cyst Apophyseal joint Cyst develops as a result of apophyseal joint degeneration
Ligamentous hypertrophy
Spinal nerve
Sciatic nerve
Synovial Cyst
Spinal nerves
G
Piriformis Syndrome
Gluteal Injection-Site Trauma
Injury of the piriformis muscle compresses the sciatic nerve Sacrum
Femur
Piriformis muscle
Sciatic nerve
POSTERIOR VIEW
H
Gluteus maximus
Obstetrical Sciatic Compression
I
Pubis
Head of fetus compresses sciatic nerve against the sacrum
Pelvic-Floor Tumors Abdominal cavity
Birth canal
Sciatic nerve
Injury varies depending on the location and agent being injected
Tumor
Soft-tissue tumor exerts pressure on the lumbosacral plexus and sciatic nerve
Head of fetus Lumbosacral plexus
Sciatic nerve
INFERIOR
Sacrum
Pelvis
Sciatic nerve
L ATERAL
VIEW
VIEW
Figure 1. Origin and Course of the Sciatic Nerve and the Main Sites of Damage. Panel A shows the normal anatomy surrounding the sciatic nerve. Panels B through I show the main types of damage.
1242
n engl j med 372;13
nejm.org
March 26, 2015
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
Sciatica
A
Lasègue’s Test
Increased pain on dorsiflexion of the patient’s foot increases sensitivity of the test Peroneal nerve Tibial nerve
Sciatic nerve
40 to 50 degrees
B
Spinal Nerve Roots under Traction
L5 nerve root
Dorsal root ganglion
Intervertebral disk L3–L4
Cauda equina L4
L5
L3
Pedicle
L2
Sacrum
Herniated disk at L4–L5 compressing L5 nerve root POSTERIOR VIEW
Figure 2. Straight-Leg-Raising Test. Panel A shows the straight-leg-raising test, with assessment for the presence or absence of Lasègue’s sign, and Panel B shows the spinal nerve roots under traction during the test.
intervertebral space, spondylolisthesis, osteomyelitis, or tumor infiltration of a vertebral body. The nature and location of disk rupture and spinal lesions, such as osteoarthritic disease and spondylolisthesis, lateral recess stenosis, and synovial cysts of the facet joint, can be accurately seen on magnetic resonance imaging (MRI) without the administration of gadolinium (Fig. 3). n engl j med 372;13
Computed tomography (CT) is performed less frequently but reveals most disk herniations and structural changes of the spine. CT and MRI each have a role in exploration of the pelvic fossa when sciatic compression is suspected in those regions. Disk bulging and related minor deformities generally do not cause compression of nerve roots and are not an explanation for persistent sciati-
nejm.org
March 26, 2015
1243
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
The
n e w e ng l a n d j o u r na l
of
m e dic i n e
findings are fibrillations and sharp waves in muscles corresponding to a single nerve root, combined with normal sensory potentials from nerves that carry that root; the latter is the result of the preganglionic site of root compression, leaving the cell body and distal axon unaffected. Sampling four or five relevant muscles including the paraspinal regions is considered adequate.9 Late responses are often included in the nerveconduction studies because they are sensitive to root compression. Electrodiagnostic confirmation that sciatica is due to a radiculopathy rather than to an undefined musculoskeletal problem is associated with a better outcome,10 but the role of EMG in sciatica has not been established and some guidelines do not require testing.
A
Nonspina l C ause s of Sci at ic a
B
Sciatica due to disk rupture and degenerative spine disease is more common than sciatica due to all nonspinal causes taken together (Table 1 and Fig. 1). Hip disease and sciatica are likely to be mistaken for each other. Piriformis Syndrome and Back-Pocket Sciatica
Figure 3. MRI Scan of a Ruptured Intervertebral Disk. Sagittal (Panel A) and axial (Panel B) T2 -weighted images show a herniated disk (arrows) at L4–L5. Compression of the right L5 nerve root can be seen as it exits the neural foramen (arrowhead).
ca. The clinician should be aware that a high rate of minor abnormalities are seen on MRI in patients without back or radicular symptoms but that disk rupture is seen in less than 1% of asymptomatic patients.8 Needle electromyography (EMG) and nerveconduction studies aid diagnosis by revealing a topographic distribution of muscular denervation corresponding to a nerve root. Denervation becomes apparent days or weeks after injury, and the appearance of changes takes longer in distal than in proximal muscles. The most specific
1244
n engl j med 372;13
The piriformis syndrome has been attributed to compression of the sciatic nerve underlying the piriformis muscle, which stabilizes and serves as an external rotator of the hip. Although the nature and frequency of the syndrome are uncertain, the most consistent features, derived from a systematic review,11 are focal mid-buttock pain, tenderness over the sciatic notch, aggravation of pain after sitting, and provocation with maneuvers that increase piriformis muscle tension such as external rotation of the hip. Pain with the straightleg-raising maneuver was reported in a metaanalysis to occur in approximately half of cases. It is presumed that an injured piriformis muscle, sometimes caused by running, stretching, or lunging, compresses the sciatic nerve, but the pathophysiology is unclear. Electrodiagnostic and imaging tests are usually normal. In the mid-20th century, as wallets expanded in capacity, sciatica (“credit-carditis”) that was cured by removal of the wallet from the back pocket became the subject of anecdotal reports.12,13 Tools, cell phones, and golf balls in back pockets and prolonged sitting on hard surfaces, including car seats, can also produce sciatica.
nejm.org
March 26, 2015
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
Sciatica
Zoster Sine Herpete
For the first few days before the eruption of shingles in a lumbar or upper sacral dermatome, herpes zoster reactivation closely simulates sciatica caused by disk rupture. Diagnosis is difficult in the intervening days and especially in the few instances in which the rash never develops.
Table 1. Causes of Sciatica and of Conditions Simulating Sciatica. Spinal causes Disk rupture with compression of L4, L5, or S1 root Osteoarthritic impingement on roots due to spondylolisthesis or foraminal or lumbar stenosis Synovial cyst of facet joint
Traumatic Injury of the Sciatic Nerve
Arachnoid (Tarlov) cyst
Traumatic injury of the sciatic nerve occurs with fractures of the pelvis or after proximal hamstring injury and with extreme stretching of the nerve. Muscle hematomas or tendinous injuries may cause severe sciatica. Posterior hip dislocation or femoral fracture impinge on the nerve; sciatica may also occur during relocation of the hip joint. Injection-site damage in the buttocks is a self-evident but infrequent cause.
Tumors of the spinal column
Gynecologic and Peripartum Causes
Zoster sine herpete
Neurofibromas on lumbar and sacral nerve roots Arachnoiditis Nonspinal causes Pelvic and gynecologic conditions, including cyclic sciatica due to endometriosis Piriformis syndrome and back-pocket sciatica Pregnancy, delivery, and prolonged time in the lithotomy position
Diabetic radiculopathy Deposits of endometrial tissue on the proximal Gluteal injection-site trauma nerve can cause recurrent cyclic, or catamenial, sciatica, more often on the right side than on the Lumbar plexitis left side.14 Large ovarian cysts and uterine enlargeVascular impingement on sciatic nerve (pseudoment associated with late pregnancy can compress aneurysm of the inferior gluteal artery) the nerve between the fetal head and the pelvic Hip fracture and dislocation or relocation; extensive pelvic fractures brim (Fig. 1). It is not unusual for a woman to have unilateral or bilateral sciatica post partum, Biceps femoris hematoma, strain, and tear regardless of whether forceps were used during Idiopathic causes in young persons delivery; sciatica ostensibly occurs after a woman has spent a prolonged time in the lithotomy position. tions may provide short-term relief for low back and sciatic pain; however, it is difficult to determine their effect on sciatica,15 and many patients C onservat i v e T r e atmen t report little relief. Orally or systemically adminof Sci at ic a istered glucocorticoids have been used to amelioSciatica resolves without treatment in the major- rate sciatica, but it is difficult to interpret their ity of cases. Numerous systematic analyses have effect.15-17 Many guidelines recommend restrictions compared various forms of therapy for sciatica, but on the use of opioids. Antiepileptic drugs, gabathe validity of their constituent studies is limit- pentin, pregabalin, antidepressant agents (e.g., tried.15-17 Many conflate treatment for back pain with cyclic agents), muscle relaxants, and pain medicathat for sciatica. Although it is difficult to derive tions have been used but with little supporting data. clinically useful information from these compiAny intervention that temporarily reduces pain lations, they do provide some perspective on the and increases participation in exercises, even if it relative value of treatments. is associated with no long-term benefit, could play The most common initial treatment is pain a role in care. The benefits of physical therapy and control by means of medication and physical various exercise regimens are difficult to detertherapy. Activity is usually self-limited in propor- mine, and the superiority of any one program has tion to the degree of discomfort, and although not been established, although most appear to be rest is often recommended, it is not better than safe. Programs include directional preference exmovement in patients who are able to remain ercises (back-movement exercises in a direction active.18 Nonsteroidal antiinflammatory medica- that causes the locus of pain to move proximally,
n engl j med 372;13 nejm.org March 26, 2015
1245
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
The
n e w e ng l a n d j o u r na l
toward the mid-back, where it is better tolerated),19 motor-control exercises (also known as specific stabilization exercises) that focus on enhancing control of the transversus abdominis and multifidus muscles, which stabilize the spine, strengthening of other core muscles, stretching, general fitness exercises, and yoga. One conventional approach involves slowly increasing mobilization of the lumbar spinal segments by means of stretching and exercise, improving posture, and strengthening the muscles that stabilize the spinal column and pelvis. It has been difficult to show that this approach accelerates recovery or prevents future injury, but some trials suggest that it is superior to rest in the acute phase of sciatica.20 Spinal manipulation for sciatica is widely used and has been studied with an assortment of designs and comparators; therefore, reviews of existing trials, most considered of low or moderate quality, draw limited conclusions.16,17,21 One recent pragmatic trial for subacute or chronic back and leg pain has suggested that spinal manipulation provides a short-term benefit when added to an exercise program.22 Lumbar traction was used in the past, particularly if low back pain was prominent, but a review of available studies, most of low and moderate quality, showed no advantage.23 Transcutaneous electrical nerve stimulation is probably ineffective.24 Acupuncture has been proposed for persistent sciatic discomfort, but its efficacy has not been shown.15,16 Chemonucleolysis of an extruded disk had positive results in some studies and systematic analyses but is infrequently used. Epidural injections of glucocorticoids are frequently administered for low back pain and related conditions. Trials have suggested an associated short-term decrease in leg pain but no decrease in the need for subsequent surgery.16,17,25 There is no clear difference in efficacy between fluoroscopically guided translaminar and transforaminal injections, but some guidelines endorse one technique over the other. In experimental studies and preliminary clinical trials, biologic agents such as tumor necrosis factor inhibitors have shown negative or limited positive effects on leg pain.26 Treatment of the piriformis syndrome involves stretching and physical therapy to enhance mobility. Injections into the muscle are sometimes administered under fluoroscopic, electromyographic, or ultrasonographic guidance; the use of glu1246
of
m e dic i n e
cocorticoids and botulinum toxin has been examined alone and in combination with physical therapy, with the suggestion that botulinum toxin may be more helpful than either placebo or injections containing a combination of lidocaine and a glucocorticoid in relieving discomfort,27 but support for any of these treatments is weak.
Surgic a l T r e atmen t of Sci at ic a C aused by Lumb a r Disk Dise a se Sciatica resolves without treatment in one third of patients within 2 weeks and in three quarters of patients within 3 months after onset.28 Nevertheless, most trials comparing surgical treatment and conservative treatment of sciatica due to lumbar disk disease favor surgery, because it results in earlier relief of pain. One representative study showed that patients who had sciatica for 6 to 12 weeks and were assigned to lumbar disk surgery had faster and more pronounced pain relief than those assigned to conservative treatment. However, 1 year after surgery, there were few differences in pain or disability between the two groups.29 The trial also showed that 39% of patients initially assigned to conservative therapy required surgery after a median of 14 weeks of pain, and in a 5-year follow-up study, an additional 7% requested surgery.30 One may conclude that surgery produces faster pain relief and accelerates improvement in mobility but that postponement of surgery to determine whether pain will abate is appropriate. It is not known whether delaying surgery in persons with weakness reduces the chance of full recovery or allows time for weakness to resolve. Surgery has been recommended in cases of large disk ruptures into the spinal canal that compress the cauda equina and cause failure of the bladder sphincter or the bowel sphincter. In a cost-effectiveness study of prolonged conservative care versus early surgery, the shortened duration of pain after diskectomy correlated with a cost savings per quality-adjusted life-year of approximately $60,000.31 In that study, 23% of the participants reverted from recovered to not recovered or vice versa at various times over a 5-year period, which indicates that sciatica can be a chronic and relapsing symptom. In another study, the configuration of the disk on MRI 1 year after the onset of symptoms did not distinguish between patients in whom sciatica had improved and those in whom it had not improved; inexpli-
n engl j med 372;13 nejm.org March 26, 2015
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
Sciatica
cably, this was true for patients who underwent treatment of low back pain in patients with and surgical treatment and for those who underwent in those without sciatica, the outcomes and the conservative treatment.32 cost of care have remained unchanged for more than a decade.35 A review of major trials with adequate data for analysis concluded that there Surgic a l Technique s was conflicting evidence on long-term benefit but Treatment of sciatica by means of decompression that surgery relieved pain more rapidly and to a of a lumbar nerve root is most likely to succeed if greater degree than did conservative therapy.36 symptoms conform to the typical clinical pattern A discursive analysis by the National Institute and imaging studies show disk rupture. Several of Health Research, in the United Kingdom, about open and percutaneous surgical approaches have the effectiveness of various strategies to manage been devised. A unilateral hemilaminotomy (re- sciatica15,37 showed some support for almost all moving parts of adjacent laminae on the side that current treatments, including disk surgery, epineeds to be decompressed) is usually adequate. dural glucocorticoids, chemonucleolysis, and alIn the past, bilateral laminectomy, a more exten- ternative therapies. However, only surgery bensive procedure, was performed, but the unilateral efitted all aspects of global effect, pain relief, and procedure is favored because it preserves tension a composite of condition-specific outcomes in and alignment between adjacent spinal segments. the short, medium, and long term. Microdiskectomy and various minimally invasive North American Spine Society guidelines38 and percutaneous techniques involving, for ex- state that diskectomy provides more effective and ample, a series of tubular retractors or an endo- more rapid symptom relief than do other treatscope, are also used to treat disk rupture. ments for symptoms that warrant surgery, alThe sequestered disk fragment or larger though less severe symptoms can be managed amounts of the intra-annular disk may be re- conservatively; patients with psychological dismoved. Fusion of adjacent spinal segments is gen- tress have poor outcomes after surgery; epidural erally not necessary if there is no mobile spondy- glucocorticoid injection provides short-term pain lolisthesis and the operation is at one level. In relief; electrodiagnostic studies have limited utilpooled analyses, there is little evidence that any ity; and for many technical aspects of surgery, technique, minimally invasive or conventional, such as lumbar fusion, there was insufficient produces better results than any other technique.33 evidence for comment. The guidelines also indiMinimally invasive approaches show a trend to- cate that there is inadequate evidence to gauge ward earlier and better pain relief, probably be- how long one can wait before surgery and still cause there is less paraspinal muscle damage, but recover from cauda equina compression or moaccording to a meta-analysis, they require longer tor deficits. British Pain Society guidelines for the treatoperating times and are associated with a higher rate of rerupture of the disk34 and therefore may ment of low back pain and radicular pain39 broadnot be appropriate in all cases. Complications of ly reflect other analyses in separating the plans for surgery are infrequent but include dural tears radicular pain from those for low back pain and with leakage of cerebrospinal fluid, as well as acknowledge that the recommendations lack predamage to the root or cauda equina. The use of cision. According to the British Pain Society, any hardware to prevent instability and chronic low indication of cauda equina compression requires back pain is often necessary in cases of spinal referral for urgent surgery. Severe radicular pain stenosis, in which multilevel operations or those that is disabling or intrusive or that prevents the involving wide resection of bone are typically patient from performing everyday tasks requires performed. MRI, as does a persistent neurologic deficit that lasts 2 weeks. MRI should be ordered by clinicians who can interpret the results; it is discourGuidel ine s a nd S ys tem at ic aged at the primary care level. R e v ie ws Despite the numerous guidelines and systematic reviews that have been published regarding the
No potential conflict of interest relevant to this article was reported. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.
n engl j med 372;13 nejm.org March 26, 2015
1247
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
Sciatica
References 1. Pearce JMS. A brief history of sciatica. 16. Luijsterburg PA, Verhagen AP, Ostelo Spinal Cord 2007;45:592-6. RW, van Os TA, Peul WC, Koes BW. Ef2. Frymoyer JW. Back pain and sciatica. fectiveness of conservative treatments for N Engl J Med 1988;318:291-300. the lumbosacral radicular syndrome: a 3. Mixter WJ, Barr JS. Rupture of the in- systematic review. Eur Spine J 2007; 16: tervertebral disk with involvement of the 881-99. spinal canal. N Engl J Med 1934;211:210-5. 17. Valat JP, Genevay S, Marty M, Rozen4. Porchet F, Wietlisbach V, Burnand B, berg S, Koes B. Sciatica. Best Pract Res Daeppen K, Villemure JG, Vader JP. Rela- Clin Rheumatol 2010;24:241-52. tionship between severity of lumbar disc 18. Vroomen PC, de Krom MC, Wilmink disease and disability scores in sciatica JT, Kester AD, Knottnerus JA. Lack of efpatients. Neurosurgery 2002;50:1253-9. fectiveness of bed rest for sciatica. N Engl 5. Lindblom K, Rexed B. Spinal nerve J Med 1999;340:418-23. injury in dorso-lateral protrusions of lum- 19. Surkitt LD, Ford JJ, Hahne AJ, Pizzari bar discs. Neurosurgery 1948;5:413-32. T, McMeeken JM. Efficacy of directional 6. Konstantinou K, Dunn KM. Sciatica: preference management for low back review of epidemiological studies and pain: a systematic review. Phys Ther 2012; prevalence estimates. Spine (Phila Pa 1976) 92:652-65. 2008;33:2464-72. 20. Kennedy DJ, Noh MY. The role of core 7. van der Windt DA, Simons E, Ripha- stabilization in lumbosacral radiculopagen II, et al. Physical examination for thy. Phys Med Rehabil Clin N Am 2011;22: lumbar radiculopathy due to disc hernia- 91-103. tion in patients with low-back pain. Co- 21. Leininger B, Bronfort G, Evans R, Rechrane Database Syst Rev 2010; 2: iter T. Spinal manipulation or mobilizaCD007431. tion for radiculopathy: a systematic re8. Jensen MC, Brant-Zawadzki MN, Obu- view. Phys Med Rehabil Clin N Am 2011; chowski N, Modic MT, Malkasian D, Ross 22:105-25. JS. Magnetic resonance imaging of the 22. Bronfort G, Hondras MA, Schulz CA, lumbar spine in people without back pain. Evans RL, Long CR, Grimm R. Spinal maN Engl J Med 1994;331:69-73. nipulation and home exercise with advice 9. Tsao BE, Levin KH, Bodner RA. Com- for subacute and chronic back-related leg parison of surgical and electrodiagnostic pain: a trial with adaptive allocation. Ann findings in single root lumbosacral ra- Intern Med 2014;161:381-91. diculopathies. Muscle Nerve 2003;27:60- 23. Wegner I, Widyahening IS, van Tulder 4. MW, et al. Traction for low-back pain with 10. Savage NJ, Fritz JM, Kircher JC, Thac- or without sciatica. Cochrane Database keray A. The prognostic value of electro- Syst Rev 2013;8:CD003010. diagnostic testing in patients with sciati- 24. Buchmuller A, Navez M, Milletre-Berca receiving physical therapy. Eur Spine J nardin M, et al. Value of TENS for relief of 2014 July 22 (Epub ahead of print). chronic low back pain with or without 11. Hopayian K, Song F, Riera R, Sam- radicular pain. Eur J Pain 2012;16:656-65. bandan S. The clinical features of the 25. Carette S, Leclaire R, Marcoux S, et al. piriformis syndrome: a systematic review. Epidural corticosteroid injections for sciEur Spine J 2010;19:2095-109. atica due to herniated nucleus pulposus. 12. Battle JD. “Credit-carditis” — a new N Engl J Med 1997;336:1634-40. clinical entity? N Engl J Med 1966;274: 26. Genevay S, Viatte S, Finckh A, Zuffer467. ey P, Balagué F, Gabay C. Adalimumab in 13. Gould N. Back-pocket sciatica. N Engl severe and acute sciatica: a multicenter, J Med 1974;290:633. randomized, double-blind, placebo-con14. Forrest JS, Brooks DL. Cyclic sciatica trolled trial. Arthritis Rheum 2010; 62: of endometriosis. JAMA 1972;222:1177-8. 2339-46. 15. Lewis RA, Williams NH, Sutton AJ, et 27. Fishman LM, Anderson C, Rosner B. al. Comparative clinical effectiveness of BOTOX and physical therapy in the treatmanagement strategies for sciatica: sys- ment of piriformis syndrome. Am J Phys tematic review and network meta-analy- Med Rehabil 2002;81:936-42. ses. Spine J 2013 October 4 (Epub ahead 28. Vroomen PC, de Krom MC, of print). Knottnerus JA. Predicting the outcome of
1248
sciatica at short-term follow-up. Br J Gen Pract 2002;52:119-23. 29. Peul WC, van Houwelingen HC, van den Hout WB, et al. Surgery versus prolonged conservative treatment for sciatica. N Engl J Med 2007;356:2245-56. 30. Lequin MB, Verbaan D, Jacobs WCH, et al. Surgery versus prolonged conservative treatment for sciatica: 5-year results of a randomised controlled trial. BMJ Open 2013;3(5):e002534. 31. van den Hout WB, Peul WC, Koes BW, Brand R, Kievit J, Thomeer RT. Prolonged conservative care versus early surgery in patients with sciatica from lumbar disc herniation: cost utility analysis alongside a randomised controlled trial. BMJ 2008; 336:1351-4. 32. el Barzouhi A, Vleggeert-Lankamp CL, Lycklama à Nijeholt GJ, et al. Magnetic resonance imaging in follow-up assessment of sciatica. N Engl J Med 2013; 368:999-1007. 33. Jacobs WC, Arts MP, van Tulder MW, et al. Surgical techniques for sciatica due to herniated disc, a systematic review. Eur Spine J 2012;21:2232-51. 34. Chang X, Chen B, Li HY, Han XB, Zhou Y, Li CQ. The safety and efficacy of minimally invasive discectomy: a metaanalysis of prospective randomised controlled trials. Int Orthop 2014;38:1225-34. 35. Mafi JN, McCarthy EP, Davis RB, Landon BE. Worsening trends in the management and treatment of back pain. JAMA Intern Med 2013;173:1573-81. 36. Jacobs WC, van Tulder M, Arts M, et al. Surgery versus conservative management of sciatica due to a lumbar herniated disc: a systematic review. Eur Spine J 2011; 20:513-22. 37. Lewis R, Williams N, Matar HE, et al. The clinical effectiveness and cost-effectiveness of management strategies for sciatica: systematic review and economic model. Health Technol Assess 2011; 15: 1-578. 38. Kreiner DS, Hwang SW, Easa JE, et al. An evidence-based clinical guideline for the diagnosis and treatment of lumbar disc herniation with radiculopathy. Spine J 2014;14:180-91. 39. Lee J, Gupta S, Price C, Baranowski AP Low back and radicular pain: a pathway for care developed by the British Pain Society. Br J Anaesth 2013;111:112-20. Copyright © 2015 Massachusetts Medical Society.
n engl j med 372;13 nejm.org March 26, 2015
The New England Journal of Medicine Downloaded from nejm.org at WEILL CORNELL MEDICAL COLLEGE LIBRARY on September 27, 2016. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
